Abstract: Aiming at the selective assembly problem for complex mechanical product with multi-objective assembly function,a method based on genetic algorithm was proposed.A comprehensive optimization model taking the assembly precision and assembly yield into consideration was established,which had combined geometric tolerance and dimensional tolerance.It could effectively reduce the geometric tolerance of the product under the premise of ensuring the assembly precision of the dimensional tolerance.A priority evaluation model based on contribution degree was proposed,which effectively improved the convergence speed of genetic algorithm.According to the characteristics of dimension chain of complex mechanical products,a kind of coding method based on spare parts was used,and mapping association matrix was established to describe the relationship between multiple tolerance items.The strength value and density information were taken into account to build the fitness function,which was taken as the evaluation rule of the individuals.The assembly of a crank and connecting rod mechanism verified that the method was feasible and effective.

Key words: selective assembly, shape and location tolerance, genetic algorithms, assembly quality, complex mechanical product

摘要： 针对复杂机械产品多质量要求下的选配问题,提出一种基于遗传算法的选择装配方法。以装配精度和装配成功率作为质量要求的评价指标,建立了综合考虑形位公差与尺寸公差的装配质量综合优化模型,能够在保证尺寸公差装配精度的前提下,有效降低产品的形位公差。同时提出了基于影响度的选配优先级评价模型,有效地提高了遗传算法的收敛速度。根据复杂机械产品尺寸链的特点,提出一种以零部件为单元的编码方式,并建立了映射关联矩阵描述多个公差项间的关联关系,综合Pareto支配强度及密集度生成适应度函数作为个体评价规则,以某发动机曲柄连杆机构的装配为例验证了该方法的可行性和有效性。

关键词: 选择装配, 形位公差, 遗传算法, 装配质量, 复杂机械产品